← All Work 03 / 06
Redesign Information Architecture Card Sorting

Placemaking Week Website Redesign

Restructuring and enhancing the desktop and mobile experience for Project for Public Spaces' annual conference — fixing a dynamic navigation system that was actively discouraging potential attendees.

Role
UX Researcher · UX Designer
Team
David Bradshaw, Devin Singh, Nicole Tominaga, Yichen Wang
Duration
14 weeks · Spring 2020
Tools
Adobe XD · Optimal Workshop
Placemaking Week Website Redesign
01 — Problem

Inconsistent Navigation & Cluttered Information

Placemaking Week — a subset of Project for Public Spaces — is a weeklong global gathering emphasizing hands-on learning and innovative social events, while leaving behind a public space legacy in host cities around the world.

The current website suffered from two compounding problems: it was text-heavy to the point of overwhelming users, and its dynamic navigation menu created a labyrinth with no clear escape. The "back" button was users' only rescue. Together, these two factors were actively discouraging potential attendees from registering.

Problem 01
Dynamic navigation with no persistent menu — users lost their place constantly and couldn't form a mental model of the site structure.
Problem 02
Landing page failed to entice — 2 of 3 observation study participants said they would not want to attend based on their first impression alone.
Problem 03
Text-heavy pages without clear hierarchy — users were unsure whether reading through content would even lead them to what they needed.

"[This is] too much text, I am not gonna read through all of this."

— Observation Study Participant
02 — Research

Mixed Methods Study

We focused on students as our primary user group — particularly interesting given that they represent only 5% of total attendees, yet are a key demographic for building long-term conference community. I consulted students from business, design, computer science, and architecture to capture a diverse range of needs.

We combined qualitative data from 15 in-person interviews and 10 usability tests with quantitative data from an online questionnaire to build a comprehensive picture of user needs.

Questionnaire Results
65% of participants had never attended a placemaking or urban design conference
Observation Study Notes
Notes from observation study sessions — conducted in-person and then adapted to remote
User Persona
Persona: Architecture Student — the target user group for this redesign
03 — Architecture

Fixing the Information Architecture

Preliminary research made it clear that fixing the IA was the single highest-priority intervention. Our high-level navigation goals were clear:

  • Consistent navigation menu across all pages
  • Simplified structure accessible to every user group
  • Intuitive labelling system that maps to user mental models

Card Sorting — 38 Cards, 11 Participants

We defined 38 cards combining previous user input, terminology from the existing site, and language used by other well-known conferences. The card sort was conducted with 11 participants using Optimal Workshop.

The sort partially failed — we ended up with 61 unique categories. Rather than a setback, this revealed the depth of labelling confusion. Users weren't confused by the content; they were confused by the language used to describe it.

Card Sorting Similarity Matrix
Similarity matrix from card sort — 61 unique categories signalled systemic labelling confusion

Tree Testing — 10 Participants, 7 Tasks

Taking card sort insights, we modified the labelling system and ran tree testing with 10 participants (5 in-person, 5 remote) using 7 short tasks designed to test navigation across key user journeys.

Final Tree
Final information architecture tree tested across 10 participants
Tree Testing Results
Tree testing results overview — directness and success rates by task
Revised Site Map
Revised site map — consistent, simplified architecture based on research findings
04 — Design

From Paper to Pixel

Every team member was responsible for sketching one section. I owned the header and footer — drawing two header iterations and three footer iterations. We then digitized paper prototypes and tested them remotely with 5 users (COVID forced the pivot from our in-person plan).

Key Feedback from Prototype Testing

  • Users expected clicking on speaker photos to reveal more information
  • Navigation menu order was counterintuitive — users expected "About" first
Paper Prototype Sketches
Initial sketches — header and footer iterations for early concept validation

Style Guide

The visual redesign aimed to retain the organization's identity while updating components for better contrast and aesthetic appeal. The redesign uses the Open Sans typeface with a responsive 12-column grid system, and a color palette that inherits legacy while improving contrast ratios.

High Fidelity Screens
Final high-fidelity designs — desktop and mobile for three primary pages
05 — Outcome

Impact & Reflections

The client appreciated how the research-driven IA improvements translated directly into cleaner navigation and a more inviting first impression. We redesigned three pages for both desktop and mobile platforms.

Learning 01
A partially failed card sort is still valuable data — 61 categories revealed labelling confusion that open-ended interviews had not surfaced as clearly.
Learning 02
Remote pivots require deliberate scaffolding. Digitizing paper prototypes quickly while maintaining test fidelity was a genuine design challenge in itself.
Learning 03
When the core problem is IA, fixing visual design first is a trap. Research-first sequencing — card sort, tree test, then wireframes — was the right call.